►
Description
Date: 10/5/2018
Presenter: Matthew Lange
Institution: University of California, Davis
West Big Data Hub
A
He
also
Co
leads
the
go
Dan
working
group
for
alignment
of
authoritative
vocabularies
of
food
and
is
designing
on
two
ontological
infrastructure
for
a
new
food
central
repository
for
the
USDA
food
composition,
databases.
He
also
teaches
database
and
knowledge
modeling
in
the
Graduate
health
informatics
program
at
UC,
Davis
School
of
Medicine
and
I'd
like
to
add
I,
was
so
fortunate
to
be
at
the
last
IC
foods
summit
and
I
believe
the
next
one
is
coming
up.
A
I,
don't
know
if
I'll
be
there,
but
others
might
be
able
to
make
it
in
November
and
what
I
really
think
is
interesting
about
what
Matthew
is
doing
is
he
provides
such
an
interesting
way
of
organizing
data
that
crosses
several
fields
of
people
who
wouldn't
usually
be
talking
to
each
other,
to
bring
new
insights?
It's
really
all
the
stuff.
We
always
talk
about
facilitating
with
infrastructure
and
great
systems,
and
so
much
of
that
starts
with
the
the
people
networking,
which
is
also
what
the
hub's
is
all
about.
So
with
that,
please
take
it
away.
B
Christine,
thank
you
for
such
a
nice
introduction.
I
really
appreciate
that.
Can
everybody
hear
me
all
right?
Yes,
terrific
and
you
can
see
my
title
slide
all
right.
Yes,
fantastic,
so
I
use,
Google,
slides
and,
and
if
you
want
to
take
a
minute
and
just
write
that
down
that
link,
then
you
have
access
to
the
slides
and,
and
actually
there
will
probably
be
a
bunch
more
so
I've
got
it.
B
Our
first
conference
that
we
had
happened
over
Election
Day
in
2016
and
yeah.
There
you
go.
That
was
that's
exactly
what
we
were
trying
to
avoid
again
and
and
and
as
as
hot
as
as
the
political
nature
of
things
are
right
now
made
the
executive
decision
that
I
don't
want
to
be
anywhere
near
an
election
day
with
a
conference
ever
again,
so
we
will
be
March
22
to
25
next
year
and
that
is
actually
going
to
be
after
a
full
week
of
other
food
conferences.
B
So
we
have
the
Agrotech
the
world
Agrotech
summit
earlier
in
the
week
and
then
future
food
tech,
SF
and
then
IC
foods
which
is
dedicated
to
building
these
internet
food.
So
there
are
be
I,
think,
nine
days
in
a
row,
three
different
conferences
for
people
who
are
interested
in
food
and
technology,
but
today
I'm
going
to
talk
to
you
a
little
bit
about
the
semantic
web
of
food
as
we
see
it
coming
and
the
Internet
of
food.
Now.
B
The
idea
here
is
that
we
can
build
vocabularies
for
food
that
span
from
molecules
to
earth
observations
that
span
in
in
in
dimension
that
way
that
span
in
discipline
from
essentially
agronomy
to
zoology,
and
it's
a
tall
order.
It's
it's
an
ocean
boiling
exercise,
but
there's
no
reason
that
each
one
of
these
disciplines
should
be
using
different
words
for
food
so
and
what
we
see
coming
in
the
future
right.
We've
already
got
lots
of
sensors
and
robotics,
and
we've
got
this
amazing
cloud
storage
abilities.
B
We
were
just
lists,
you
know,
hearing
about
NSF
and
petabytes
of
storage
and
things
like
that
and
and
the
new
blockchain
and
distributed
Ledger's
and,
of
course,
now
in
the
last
half
a
dozen
years
machine
learning
AI
has
come
on
strong
for
making
sense
of
these
massive
amounts
of
data,
and
it's
pretty
easy
that
we
want
to
make
a
better
planet.
We
want
to
improve
the
health
of
people
and
what
I
think
is
particularly
important
and
and
and
I
our
pond.
B
This,
when
I
teach
medical
students
and
in
particular
people
in
the
medical
profession,
is
that
food
is
really
about
performance
and
delight
we
don't.
Sometimes
we
eat
food
to
get
better,
but
generally
we're
eating
food
to
nourish
ourselves
into
to
improve
our
quality
of
life,
as
it
is
now
not
to
fix
something,
that's
gone
wrong,
so
it
doesn't
exactly
fit
into
the
model
of
our
healthcare
system,
which,
I
think
is
actually
miss
named.
B
The
amount
of
data
on
the
current
World,
Wide
Web,
so
by
2020
IOT,
will
be
bigger
than
WW.
Now,
if
you
think
about
food
being
influx
through
many
of
these
devices,
and
and
you
know
we
just
heard
Walmart
wants
to
track
food-
you
know
making
their
suppliers
use.
Blockchain
and
tracking
food
through
the
supply
chain
and
I'll
talk
a
little
bit
more
about
that
in
a
minute
and
and
all
this
information
coming
from
soil,
sensors
and
weather
sensors-
and
you
know
we
have
earth
observation
data.
B
So
if
you
think
about
the
amount
of
data
about
food
and
then
you
consider
omics
about
food,
what
what
am
I
eating?
What
is
the
metabolomic
of
the
plant,
and
so
it's
pretty
clear
that
that
food,
since
we
all
need
every
day
influx
through
the
internet
of
things,
that
the
iof
will
actually
be
orders
of
magnitude
larger
than
the
IOT,
which
is
a
this
kind
of
mind-boggling
to
think
about
the
size
of
what
we're
talking
about
here.
B
So
our
vision
here
with
IC
foods
is
really
to
create
a
global
technological
infrastructure.
That's
capable
of
collapsing
the
time
that
it
takes
to
translate
scientific
knowledge
into
application
right.
So
this
idea
that
we
we
produce
a
bunch
of
data
and
then
we
write
it
up
in
a
paper
and
then
some
comes
along
and
uses
natural
language
parsing
to
parse
the
meaning
of
the
paper,
and
then
it
turns
that
into
data.
That's
a
little
silly.
We
should.
B
So
we're
we're
improving
the
food
supply
here
by
enabling
people
to
compete
on
the
same
things
that
we
value
for
society
and
for
individuals
to
be
more
traceable
or
transparent
and
ultimately
to
produce
better
food,
healthier
food,
more
authentic,
more
sustainably
produced.
So
as
when
we're
traceable
and
transparent,
we
are
gonna,
be
able
to
make
decisions
based
on
how
that
corn
was
grown.
Was
the
cow.
You
know
massage
three
times
a
week
in
order
to
get
the
steak
or
or
whatever
it
is
that
you
value.
B
Maybe
you
don't
care
about
pesticides,
but
you
care
about
phosphates
in
this
soil.
What
what?
How
was?
How
are
these
things
grown?
And
ultimately
this
means
that
there's
going
to
be
all
kinds
of
new
players
and
food
systems,
so
we're
looking
at
really
democratizing
food
systems
and
having
the
same
effect
on
players
in
food
as
as
you
can
imagine
you
know,
the
internet
has
had
on
on
commerce
in
general.
B
So
when
we
think
about
food
and
the
value
propositions
for
food,
it's
really
kind
of
a
no-brainer.
Yet
we've
been
ignoring
it
all
this
time,
half
of
all
greenhouse
gases
are
related
to
food,
food,
transportation,
food
production,
food
processing,
there's
just
a
lot
going
on
there
right,
so
solving
for
food
is
also
solving
for
greenhouse
gas
emissions,
and
it's
not
just
you
know
a
bunch
of
left-wing
scientists
and
anestine
by
the
Chinese.
B
There's
there's
a
real
cost
of
doing
business
associated
with
this,
so
Swiss
reinsurance,
the
largest
reinsurance
and
in
the
world,
has
come
out
and
said.
If
you
don't
have
a
climate
solution
in
place
for
your
supply
chain,
your
insurance
is
going
to
cost.
You
more
Moody's
has
said
your
credit
is
going
to
go
down.
If
you
don't
have
supply
chain
solutions
in
place.
B
Agriculture
is
the
largest
market
for
drones
and
we
all
carry
cell
phones
and
are
able
to
track
our
health
with
cell
phones,
so
we're
producing
all
this
data
from
sensors
and
we
can
be
using
this
to
improve
our
health
and
food
supply.
So
the
example
that
I
often
give
is
that
every
web
browser
understands
HTML
hypertext
markup
language.
What
would
it
look
like
if
we
had
a
hyper
food,
markup
language?
B
Of
course,
those
of
you
who
are
in
the
language
and
software
space
realized
that
I'm,
probably
not
talking
about
markup
language
I'm
talking
about
ontology,
but
the
idea
is
that
I
can
annotate
Tomatoes
and
not
just
any
tomato.
This
tomato
right
here
in
a
time
and
place
with
a
level
of
ripeness
every
aspect
about
that
plant,
because
I've
been
monitoring
how
it
grows
with
all
kinds
of
IOT
devices.
B
So,
like
I,
said
we're
talking
about
ontologies,
which
is
in
the
top
right
here,
highlighted
in
terms
of
complexity
and
cost
ontology
czar,
perhaps
the
most
expensive
type
of
vocabulary
that
we
can
use
to
both
annotate
and
enable
semantic
technologies.
But
there
there
are
the
most
semantically
precise.
So
we've
developed
a
model
for
a
food
phenotype
ontology.
There
does
not
currently
exist
a
model
for
characterizing
all
foods
and
at
a
class
level.
So
what
we've
said
is
look
every
every
food
has
a
biological
component.
B
You
know
from
a
milk
fat
globule
membrane
to
a
to
a
pork
shoulder
they've.
Those
biological
components
contain
chemical
components
and
there
may
be
additional
structural
components
like
a
sheet
of
pasta,
and
these
all
have
properties,
biological
properties.
These
could
be
bioactive
properties
when
you
consume
specific
yoghurt
or
chemical
properties
pas
or
what-have-you
physical
properties
like
how
thick
is
that
sauce,
really
Jackel
properties,
all
viscoelastic
properties
and
then
there's
this
dependant
class
down
here
called
organoleptic
properties,
and
for
those
of
you
who
aren't
food
scientists.
B
This
is
the
smell
and
the
aroma
taste
even
sound
and
sight
of
food,
we're
working
with
a
very
large
flavor
company
right
now
to
build
out
standardized
terminologies
for
these
organ
elliptic
properties
and
that's
going
to
enable
all
kinds
of
predictability
on
flavors-
and
this
is
just
a
quick
example
of
going
in
a
more
detail
than
what
the
USDA
currently
has
as
its
nutrient
database
and
food
composition
database.
These
are
oligosaccharides
these
1
1
0
1.
A
and
all
of
these.
B
These
are
labels
of
oligosaccharides,
which
are
medium
chain,
carbohydrates
that
are
found
in
human
milk
and
are
known
to
grow
specific
gut
bacteria
to
make
the
gut
to
make
the
infant
gut
healthier,
and
these
are
all
the
cow
mares
donkeys.
These
are
all
the
animal
milks
that
contain
similar
oligosaccharides,
so
that,
if
someone
couldn't
breastfeed,
we
could
conceivably
then
create
from
a
variety
of
other
animal
milks,
specific
blends.
That
would
be
precise
milk
that
might
feed
that
baby
better.
Those
of
you
who
are
not
familiar
with
ontology
z'.
B
This
is
a
quick
look
at
the
gene,
ontology
growing,
there's,
3
root
classes,
the
biological
process,
the
molecular
function
and
the
cellular
component,
so
I
made
this
I
compiled
a
list
of
the
ontology
and
then
genes
that
were
annotated
against
the
ontology
and
through
time,
so
the
purple
guy,
that's
floating
around.
Those
are
all
different
databases
where
the
gene
was
associated
with
the
ontology
terms,
and
this
is
kind
of
a
cool
visualization,
because
what
you
get
is
this
ability
to
watch
our
collective
human
knowledge
grow
through
time,
so
I
won't
belabor.
B
That
point,
just
just
one
real
quick
sort
of
mention
here
is
that
you
can
watch.
So
if
we
have
these
like
explosions
of
knowledge
that
happen
when
people
make
there
you
go
so
this
is
all
immune
system
stuff
happening
in
the
biological
process
and
in
specific
places
in
the
extracellular
region
and
then
the
action,
the
specific
molecular
action
that
those
things
so
it
meant
whoops.
Imagine
now
we
could
do
that
for
for
for
food
and
nutrients
within
foods
as
well.
B
B
Back
and
we
can
annotate
what
I've
got
in
quotes
your
traditional
data
about
food
that
could
be
economic
data
about
food
that
could
be
sensor,
information
about
food
and
we
can
lay
all
of
this
annotated
data
down
into
a
distributed
ledger
or
a
blockchain,
and
and
all
of
this
together
comprises
the
the
IOF
and
then
this
idea
of
the
Semantic
Web
of
food.
If
we
were
to
drill
into
that
where
we
would
get
vocabularies,
we've
got
indexing
we've.
B
So
if
you,
if
you've
ever
done
any
AG
research,
you
know
that
there
is
the
calving
thesaurus,
there's
also
the
Agricultural
vocabulary
service
national
agricultural
thesaurus.
So
we
have
these
indexing
vocabularies.
We
have
vast
quantities
of
economic
and
trade
data
about
food,
socio-cultural
data,
people,
tagging
food
in
social
networks
and
things
like
that.
Pictures
of
food
on
Instagram
now
outnumber
pictures
of
cats,
which
I
think
is
quite
telling
and
and
then
we've
got
this
observation
sensor.
B
So
this
idea
that
if,
as
we
model,
this
we've
got
the
food
phenotypes,
which
I
I
mentioned
earlier
in
the
root
classes,
and
then
we've
got
the
processes
that
interact
with
food
phenotypes
the
production,
transformation,
distribution
and
consumption
processes
that
interact
and
make
those
food
phenotypes
or
or
or
consume
those
from
phenotypes.
Then
we
have
the
equipment
and
agents.
B
This
means
people
as
well
that
engage
in
processes
that
alter
food
phenotypes,
and
then
we
have
essentially
issues
of
the
people
or
agents
that
are
engaging
in
the
processes
that
are
altering
the
food
phenotypes
and
we've
we've
begun
to
develop
ontology
at
each
one
of
these
levels
and
using
interoperable
terms
between
them.
All
so
at
I
see
foods.
We've
we've
built
working
groups
for
building
on
topologies
in
each
one
of
these
areas
and
I'm
just
going
to
move
quickly.
Now
Christine
mentioned
that
we
lead
the
international
working
group
for
go
Dan
about
once
a
month.
B
We
have
calls
with
folks
from
the
USDA
the
FDA
FSA,
which
is
the
European
Food
Safety
Authority,
the
UN
Food
and
Agriculture
Organization,
the
w-h-o
and
a
variety
of
academics.
To
begin
to
align
these
vocabularies,
we
have
an
RCN
from
NSF
for
building
smart
food,
so
it's
just
getting
together
both
the
stakeholders
of
people
who
are
interested
in
data
and
the
people
who
produce
data
to
answer
the
questions.
What
do
we
eat
and
where
does
it
come
from?
What
do
we
produce
and
where
does
it
go
and
what
gets
wasted
along
the
way?
B
B
This
is
the
food
composition
database
that
pretty
much
all
food
manufacturers
in
the
United
States
and
actually
around
the
world,
use
to
talk
about
food
components
and
reason
over
them
and
we're
also
working
with
the
Cooperative
Extension
to
build
ontology
around
expertise
of
people
who
work
in
and
around
the
food
system,
and
this
was
a
poster
that
was
at
the
recent
international
conference
for
biomedical
ontologies
and
we'll
be
turning
this
into
a
full-blown
paper.
Very
soon.
We
have
a
another
project.
B
I
think
this
is
my
last
four
second,
the
last
slide
here,
where
we
are
looking
at
essentially
engineering,
a
social
movement
to
capture
the
wisdom
that
our
elders
have
about
food,
we're
losing
it
we're
losing
it
faster
I
would
argue
than
we're
losing
the
rainforest.
This
wisdom
is
dying
off
with
our
elders
and
it's
quite
a
shame
so
we'll
be
announcing
this
at
the
next
I
see
foods
conference,
basically
ten
prizes
for
ten
thousand
dollars
each
to
make
a
video
of
a
an
elder
who's.
B
Two
generations,
older
than
you
I'll,
save
the
link
for
the
the
tools
there.
You
can
go
and
explore
it.
This
is
our
team
and
our
partners
just
some
of
them
and
and
then
a
quick
look
at
their
and
I'll
stop
here
and-
and
if
there's
you
know
other
questions
after
we
can,
we
can
do
that.
Okay,
thanks
yeah.
A
B
Had
a
question:
if
that's
okay,
ontology
czar,
notoriously
difficult
to
maintain
and
carry
and
over
time,
keep
fresh
and
they'll
ontology
that
you
described
is
scarier
than
any
ontology
I've
ever
seen
so
you're
exactly
right,
so
I'm
gonna,
I'm
gonna
share
my
screen
again.
Really
quick
right!
Hey
did
you
already
take
the
controls.
B
A
A
B
So
so
you
know,
I
got
the
kind
of
hacker
mentality.
You
know
hackers
tend
to
come
into
something
and
say
this
sucks
and
every
everything
sucks,
right
and
and
the
goal
is
to
make
it
suck
less
and
so
I'm
gonna
tell
you
that
Protege
as
an
ontology,
editor,
sucks
and
and
the
current
way
that
we
track
ontology,
x'
and
and
update
them
is
so
complicated
and
heavy,
and
so
we
are
working
on
and
actually
will
be,
putting
in
an
NSF
grant
and
and
and
I
think
that
this
is
actually
going
to
align.
B
With
with
what
Gabe's
going
to
tell
you
about,
because
these
ontologies
can
link
to
business
use
cases,
but
but
suffice
it
to
say
that
you
can
you
can
drill
through
these
concepts.
This
is
the
food
on
ontology
that
I'm
a
co-author
on
and
so
I.
Can
you
know
drill
into
various
aspects
of
these
things?
What
we
don't
have
yet
is
is,
is
a
nice
way
to
to.
B
Essentially,
you
know
lasso
these
things
and
then
edit
that
area,
but
what
we've
developed
is
a
way
for
integrating
ontology
I
like
to
fight
to
say,
I,
want
to
link
Kevi
ontologies
here
and
I
want
to
link
this
food
antioxidant
with
this.
You
know
some
additive
over
here
that
might
be
an
acidifier
and
then
I
can
say
how
they're
related
to
each
other.
B
So
this
this
is
easy.
This
is
all
browser
based.
It's
super
easy.
The
idea
is
that
you'll
be
able
to
do
this
with
other
people
in
a
browser.
You
don't
need
to
be
a
PhD
on
tala,
just
to
do
this
stuff
that
you
could
actually
be
a
business
person,
and
you
could
you
know
so
we
don't
know
what
exactly
the
hierarchy
looks
like
for
for
doing
all
of
this,
but
we
think
that
you're,
you
know
that
there's
the
potential
here
to
lower
the
barrier,
to
entry
and
and
to
divide
and
conquer.
B
I'll
avoid
95
following
questions
that
this
is
really
fast.
No
I
know
it's
a
it's
an
ocean
boiling
exercise,
and
really
you
know
it's
not
gonna.
It's
we're.
Not
we're
not
going
to
do
this
ourselves
right.
So
that's
that's.
What
I
seafoods
the
consortium
is
all
about
is
about
bringing
people
together
and
solving
this
problem.
We're
gonna
boil
the
ocean
with
you
know:
10
million
candles,
I
guess
right
so
I
know.
There's
you.
B
A
Well,
one
of
the
reasons
we
invite
people
to
present
in
this
working
group
is
one
to
connect.
People,
and
you
know,
I
mean
on
top
of
just
having
you
eyes
and
functionality
that
makes
my
heart
sing
and
I.
Don't
use
those
words
lightly
or
often
I.
Just
think
this
is
such
an
interesting
project
and
I'm
glad
to
hear
there's
lots
of
interest.
I
encourage
you
to
you
know,
put
icy
foods
on
your
calendar
and
to
follow
up
with
Matthew
in
the
interim.